In the eample of pulearning,
method 1:
res = bc.oob_decision_function_[:,1]
prdict_proba = np.array([i for i in res if not np.isnan(i)])
method2 :
y_pred_prob = bagging_model.predict_proba(train_x)
The result of two method are different, which should i use to predict? I know the meaning of out of bagging. I do not know the difference of two method.